ac4 ext3 recovery failure

ac4 ext3 recovery failure

[Greg Louis]

Rebooting to try 2.4.7-ac4, I had Xfree86 crash on exit and hang the
machine (it does that once a month or so; this notebook gets booted
quite often).  After fscking the root and another ext2 partition, the
system got to the big ext3 partition and just went dead.  No message,
no disk activity, no keyboard response.  I powered down and rebooted
2.4.7-ac3 patched with ext3-2.4-0.9.5-247ac3, and that came up ok and
recovered the journalled partition.

-- 
| G r e g  L o u i s          | gpg public key:      |
|   http://www.bgl.nu/~glouis |   finger greg@bgl.nu |

Re: ac4 ext3 recovery failure

[Andrew Morton]

Greg Louis wrote:
> 
> Rebooting to try 2.4.7-ac4, I had Xfree86 crash on exit and hang the
> machine (it does that once a month or so; this notebook gets booted
> quite often).  After fscking the root and another ext2 partition, the
> system got to the big ext3 partition and just went dead.  No message,
> no disk activity, no keyboard response.  I powered down and rebooted
> 2.4.7-ac3 patched with ext3-2.4-0.9.5-247ac3, and that came up ok and
> recovered the journalled partition.

Do you think this happened during the e2fsck run, or during the
actual mount?

-

Re: ac4 ext3 recovery failure

[Alexander Viro]

On Fri, 3 Aug 2001, Andrew Morton wrote:

> Do you think this happened during the e2fsck run, or during the
> actual mount?

actual mount. s/fsync_dev/fsync_no_super/ in fs/jbd/*. Already fixed
in Alan's tree.

BTW, code around the place in question looks somewhat fishy - looks
like you have a lot of stuff assuming that journal and fs are on the
same device.

Re: ac4 ext3 recovery failure

[Andrew Morton]

Alexander Viro wrote:
> 
> On Fri, 3 Aug 2001, Andrew Morton wrote:
> 
> > Do you think this happened during the e2fsck run, or during the
> > actual mount?
> 
> actual mount. s/fsync_dev/fsync_no_super/ in fs/jbd/*. Already fixed
> in Alan's tree.

Ah.  Thanks.  Linus doesn't seem to have fsync_no_super(), so some
compatibility hacks will be needed there.

Why does fsync_no_super() exist?  I assume some locking changes
somewhere?

> BTW, code around the place in question looks somewhat fishy - looks
> like you have a lot of stuff assuming that journal and fs are on the
> same device.

Yes - this was fixed up when we were testing the external journal
support for 0.9.5.  -ac4 has 0.9.3+bits.

-

Re: ac4 ext3 recovery failure

[Alexander Viro]

On Fri, 3 Aug 2001, Andrew Morton wrote:

> Ah.  Thanks.  Linus doesn't seem to have fsync_no_super(), so some
> compatibility hacks will be needed there.

Merge with fs/super.c fixes (happens in -ac)

> Why does fsync_no_super() exist?  I assume some locking changes
> somewhere?

It's fsync_dev() without touching fs structures.

fsync_super() - we hold a superblock, sync its in-core structures and
sync the device
fsync_dev() - we have a device, find superblock, grab it and sync (see above).
We need exclusion against read_super() and umount here - code in Linus'
tree is oopsable exactly because it lacks that.
fsync_no_super() - we have a device, sync it without even looking for fs
in-core stuff.

Notice that fsync_dev() in the wrong moment can seriously screw vanilla
2.4. In -ac4 it got the required exclusion, but that means the need to
use the fsync_super() or fsync_no_super() if you are in the exclusion
area. That had been done to the stuff common with Linus' tree, so there's
not much to fix. AFAICS jbd/recovery.c is the only place not covered.